Adam Smith
Managing Director
Bellrock Technologies
Klarna has been in the news after stopping its long-term partnerships with Salesforce and Workday – reportedly because they have replaced the SaaS with new systems, they built using code created by Ai. There is a lot to unpack in that sentence, but ultimately what this has prompted is the underlying question…
Is SaaS like Concerto dead, now Ai can “whip up code on the cheap”?
Now we need to dig a firstly a bit deeper into Klarna story to get the full picture.
Firstly, it’s essential to point out that they haven’t fully built replicable software using generative Ai. From looking at previous Klarna press releases it appears they have moved from Salesforce to Freshworks and from Workday to DEEL used Ai to automate process they have lost in the transition.
This transition appears to have been over many years, and the timing and nature of the release appears to be opportunistic from a brand awareness perspective, rather than a real assumption that all software is replaceable with Ai:
- Klarna is preparing to go public, so cutting costs quickly will be key to a high share price.
- Cutting cost, particularly headcount is a better story if anchored in Ai rather than reducing bottom line to float at a higher price.
- Klarna wants to be seen as a leader in Ai.
- Klarna wants the free marketing from all the buzz
Has Ai, as a replacement for SaaS, worked already?
It’s also worth noting that they are not the only company to try this strategy. Companies, much bigger than Klarna, have tried to replicate SaaS products in this way, but after only a short period, have turned back to the original vendors.
But that being said, there are lessons for SaaS companies coming out of this story.
The question for Concerto now is what parts of an IWMS can be improved or replaced by Ai? And that is very much what our dev team are analysing at the minute.
Let’s talk about Ai and its potential
Firstly, game changing technology come and go. Think about how much organisations felt they needed to leverage Blockchain when in reality it didn’t replace relational databases in day-to-day operations. Its use is much niche.
Ai feels much more revolutionary, and much more applicable for all businesses. However, its downfall is very much its reliance on data. Reaching the “holy grail” of predictive maintenance is a great goal, but most organisations are far from being able to fully leverage that.
Most business I speak to need to use Ai to improve data quality today, cleanse your data set etc; and then worry about the predictive clever stuff at a much later date.
Let’s think about data a bit more
Ai thrives on high-quality, structured data. We have built and designed Concerto to aid that process. That’s why Bellrock Technologies invested in the acquisition of Mobiess, to close the data loop on field engineers, and create a structured way of capturing data straight into Concerto.
You can’t even think about Ai until you have completed serious legwork on your data. Strong taxonomy is essential, and it requires an investment in time and resources to get it right.
So how do you get data right?
- Curation and preprocessing of data to improve data quality. Does your data have duplicates, outliers and irrelevant data points?
- Inclusive data. You need to make sure your data has no gaps or the output will be flawed.
- Constant improvement. Create a continuous improvement loop, looking at data and reviewing the output of Ai for flaws.
How will Ai be used in the world of IWMS?
In Concerto we are developing a Natural Language Chat App. So, you can ask questions and interrogate your data using plain language. For example, “How many buildings don’t have compliant fire risk assessments?”
This makes data and analysis much more accessible.
Predictive Maintenance as a Service (PdMaaS) is also big feature of our dev work. This cloud-based service is revolutionising maintenance practices, giving even smaller organisations access to predictive tools without huge upfront investments.
The team are also keeping an eye on trends like edge computing and Internet of Things. This will drive real time data processing in Concerto and speed up decision making and reduce cloud dependency. And speed can really drive down costs and most importantly reduce risk in the world of maintenance.
Could Ai ultimately replace SaaS?
There is a chance in the long term it may replace SaaS, however as a Chief Technology Officer it’s not a risk I am particularly concerned about. Instead, I believe it will 100% change it fundamentally. And that’s the bit that as a business leader I am focussing on. When change happens those that survive are those that think completely outside the box, even if that is scary, or undermines some core fundamentals of your business. You want to be like Netflix and shift from postal DVD rentals to online streaming; rather than Blockbuster and remain stubborn in your initial offering.
Future software will be built with Ai as the interface that consumers use. And it can be difficult to imagine what that might look like, but that’s what the Bellrock Technologies research lab does. It’s deciding what our software will be in ten years’ time.
And Ai will also add value to us, as software developers, as well as our customers using our software. Ai can help with some coding. There are a lot of repeatable coding across SaaS systems, for example SSO that could easily be recreated in Ai, and therefore save hours upon hours of simple coding. However, as systems get more complex, and bespoke, then it will be difficult to replace the skill and talent coders bring to software, and the complexities of issues they rectify.
And finally, I think as Ai is introduced more heavily into software it will impact pricing models. And rather than moving from a user volume licence, it might look at consumption/usage; or possibly even outcome based.
It’s an interesting time to be part of the SaaS industry. I have been at the forefront of enterprise SaaS for over 20 years, and embracing change and innovation should only ever be seen as an opportunity to improve what you are delivering, rather than something to be feared.
